Defining Error in Anatomic Pathology

1
Defining Error in Anatomic Pathology
February 26, 2005
USCAP 05

ADASP
Ronald L. Sirota, M.D.
2
Disclosure

• Staff Pathologist, Advocate Lutheran General
  Hospital, Park Ridge, IL
• Nothing to disclose

3
Objectives

• Why are we talking about error?
• Defining error in anatomic pathology
• Error at the practitioner level

4
Tipping Point
1999
5
IOM Report on Medical Error

• To Err Is Human Building a Safer Health Care
  System-Kohn et al
• IOM report on Medical Error in America
• Written by 38 people in two subcommittees
• Published December 1, 1999

6
Medical Error Statistics

• Colorado and Utah study
• AEs in 2.9 of hospitalizations
• 8.8 of AEs led to death
• Extrapolated to country-44,000 deaths annually
• New York study
• AEs in 3.7 of hospitalizations
• 13.6 of AEs led to death
• Extrapolated to country-98,000 deaths annually
• Eighth leading cause of death in USA

7
Comparative Statistics

• Annual deaths from motor vehicle accidents-43,458
• Annual deaths from breast cancer-44,297
• Annual deaths from AIDS-16,516
• Annual deaths related to worker safety
  issues-7,000

8
(No Transcript)
9
Dollar Cost of Medical Error

• Total costs per year
• Between 17 billion and 29 billion
• Half directly attributable to health care costs
• Other half combination of lost productivity, lost
  income and disability

10
Goals of the IOM Report

• Break cycle of inaction concerning patient
  safety and medical error
• Enhance knowledge base about medical error
• Break down legal and cultural barriers which
  impede safety improvement
• Shift from blaming individuals for past errors to
  a focus on preventing future errors by designing
  safety into the health care system

11
Goals of the IOM Report

• Improve knowledge base about medical error and
  safety improvement
• Develop tools to improve safety
• Cause significant monetary investments to achieve
  these goals

12
Specific Strategies

• Create accountability throughout the healthcare
  system
• Raise standards and expectations for improvements
  in safety through the actions of oversight
  organizations, group purchasers and professional
  groups
• Regulatory bodies
• Professional societies
• Consumer groups
• Employers

13
IOM Mandate

• Threshold improvement in quality

• 50 reduction in medical error in five years

14

• Whats all this error stuff got to do with us
  ???????????????

15
(No Transcript)
16

• Health Mailbox Wall Street Journal
• Columnist Tara Parker-Pope answers readers'
  questions. February 8, 2005 Page D7
• Q Can you provide a reference for your statement
  that pathology reports in cancer cases have
  errors from 1 to 20 of the time? Is this
  someone's best guess, or is it documented by
  research?
• --D.O.
• A There is growing evidence that patients should
  always seek a second opinion on lab work when
  cancer is suspected or diagnosed, but I have been
  surprised how many skeptical doctors have written
  to me asking me for the specific studies
  supporting this. The risk of mistakes varies
  depending on the body part and type of cancer..

17

• The most-cited research comes from Johns Hopkins
  School of Medicine in Baltimore, which has shown
  that about 1.4 of pathology cases involve
  serious errors, such as diagnosing cancer when a
  tumor is actually benign or diagnosing the wrong
  type of cancer.
• In the same study, biopsies involving the female
  reproductive tract were particularly prone to
  mistakes, with an error rate of 5.1. Skin cancer
  had a pathology error rate of 2.9. In the case
  of prostate cancer, about 20 of the time
  mistakes had been made in staging and grading,
  which tell doctors how aggressive or advanced the
  cancer.
• This summer, researchers from Dana Farber Cancer
  Institute in Boston also noted a significantly
  high error rate in prostate-cancer pathology. The
  report, published in Urologic Oncology, reviewed
  602 cases. When the second opinion disagreed with
  the first, a third opinion was sought.

• The report found that the review changed the
  Gleason score -- which is the score used to
  assess the seriousness of prostate cancer -- by
  at least one point in 44 of cases. Most of the
  time, the review found the patient had more
  serious disease than originally thought. For
  about 10 of the men, the review would have
  significantly changed the treatment offered.
• A December 2002 Northwestern University study of
  340 breast-cancer patients altered lumpectomy or
  mastectomy plans for 8 of the women. And this
  month, the European Journal of Surgical Oncology
  reported that in a review of 66 cases of thyroid
  cancer at St. James University Hospital in the
  United Kingdom, the diagnosis changed in 18 of
  the cases.

WSJ, Feb 8, 2005
18

• This summer, researchers from Dana Farber Cancer
  Institute in Boston also noted a significantly
  high error rate in prostate-cancer pathology. The
  report, published in Urologic Oncology, reviewed
  602 cases. When the second opinion disagreed with
  the first, a third opinion was sought.
• The report found that the review changed the
  Gleason score -- which is the score used to
  assess the seriousness of prostate cancer -- by
  at least one point in 44 of cases. Most of the
  time, the review found the patient had more
  serious disease than originally thought. For
  about 10 of the men, the review would have
  significantly changed the treatment offered.
• A December 2002 Northwestern University study of
  340 breast-cancer patients altered lumpectomy or
  mastectomy plans for 8 of the women. And this
  month, the European Journal of Surgical Oncology
  reported that in a review of 66 cases of thyroid
  cancer at St. James University Hospital in the
  United Kingdom, the diagnosis changed in 18 of
  the cases.

19
Defining Error
20
Error

• Generic term to encompass all those occasions in
  which a planned sequence of mental or physical
  activities fails to achieve its intended outcome

21
Intended Outcomes in AP

• An accurate diagnosis
• A precise diagnosis

22
Intended Outcomes in AP

• An accurate diagnosis
• Accuracy
• The actual truth, scientifically validated, based
  on gold standards

23
Gold Standards in AP?

• Morphology?
• Immunohistochemistry?
• Cytogenetic studies?
• Banded karyotype
• FISH
• Molecular studies?

24
Intended Outcomes in AP

• An accurate diagnosis
• Accuracy
• The actual truth, scientifically validated, based
  on gold standards
• A precise diagnosis
• Precision
• A measurement of variation
• SDs, CVs, etc
• Low inter-observer variation

25
Types of Variation

• Acceptable
• No impact on therapy that can alter prognosis
• No impact on prognostication, exclusive of
  therapy
• Unacceptable
• Has major impact on therapy, which can affect
  prognosis
• Has major impact on prognostication, exclusive of
  therapy

26
Error in Anatomic Pathology

• Diagnostic variation
• Major error (acceptable error)
• Error which has a major impact on therapy that
  can alter prognosis, or
• Error which has a major impact on
  prognostication, exclusive of therapy
• Minor error (unacceptable error)
• Error which does not have a major impact on
  therapy that can alter prognosis, or
• Error which does not have a major impact on
  prognostication, exclusive of therapy

27
Error

• Not a reality, but a judgment
• A ruling made on human performance
• Always assigned after outcome is known
• Almost always affected by hindsight bias

28
Hindsight Bias

• The knew it all along effect whereby observers
  of past events exaggerate what other people
  should have been able to anticipate in foresight
• Bias induced by outcome knowledge
• Overestimation of what could have been known had
  knowledge of the outcome not been possessed

29
Hindsight Bias Results

• Distortions of the truth
• Over exaggeration of situational certainty
• Abnormalities become obvious
• Underestimation of prospective difficulties
• Rational judgements about performance are
  corrupted
• Blame is wrongly assigned

30

• Determinants of Inter-Observer Variation in
  Anatomic Pathology

31
Determinants of Inter-Observer Variation

• Taxonomic or nosological systems
• Evidence based?
• How reproducible?
• Experts
• General practitioner

32
Human factors

• Knowledge problems
• Gaps or lapses
• Inert or inaccessible
• Buggy or imperfect
• Cognitive factors
• Attention
• Biases
• Heuristics
• Memory
• Error precursors

33

• Error at the practice level of anatomic pathology

34
Practice level (individual practitioner level)

• Pre-analytic phase
• Analytic phase
• Post-analytic phase

35
Error in Anatomic Pathology Pre-Analytic

• Clinician error
• Wrong procedures, wrong specimen, wrong test,
  wrong or inadequate clinical information.
  (wrong)x
• Specimen delivery
• Accessioning error
• Wrong or improper labeling
• Wrong specimen
• Wrong demographic information
• Incorrect specimen handling
• Wrong studies ordered
• Correct studies omitted
• Histology error

36
Error in Anatomic Pathology Analytic

• Frozen section error
• Incorrectly ordered and performed
• Specimen mix-ups
• Poor sections
• Incorrect sampling
• Poor cognitive formulation
• Wrong problem addressed
• Correct problem overlooked

37
Error in Anatomic Pathology Analytic

• Gross room error
• Specimen mix-ups
• Poor gross exam
• Pertinent information corrupted (inaccurate
  description)
• Pertinent information omitted ( e.g.
  measurements)
• Wrong section taken
• Correct sections omitted
• Improper studies ordered or proper studies
  omitted

38
Error in Anatomic Pathology Analytic

• Slide mix-ups
• Final diagnostic error poor cognitive
  formulation
• Misdiagnosis
• Omission of diagnosis
• Unintelligible or incorrect report
• Format
• Wording
• Typing mistakes

39
Error in Anatomic Pathology Post-Analytic

• Lack of timeliness
• Improper delivery
• Wrong clinician
• Wrong office
• Misunderstood report